24 lines
649 B
Markdown
24 lines
649 B
Markdown
|
---
|
||
|
|
base_model: bibocat/qwen3-ner-sft-4epochs-100k-merged
|
||
|
|
library_name: transformers
|
||
|
|
pipeline_tag: text-generation
|
||
|
|
language:
|
||
|
|
- vi
|
||
|
|
tags:
|
||
|
|
- qwen3
|
||
|
|
- grpo
|
||
|
|
- medical
|
||
|
|
- named-entity-recognition
|
||
|
|
- vietnamese
|
||
|
|
---
|
||
|
|
|
||
|
|
# Qwen3 Vietnamese Medical NER — GRPO v2
|
||
|
|
|
||
|
|
BF16 merged checkpoint from the Qwen3 Vietnamese medical SFT model and the
|
||
|
|
`models/qwen3-ner-grpo-v2` GRPO v2 LoRA adapter. The adapter was trained for
|
||
|
|
one epoch on `data_train/grpo_v2/full_train.jsonl` with the span-aware v2
|
||
|
|
reward.
|
||
|
|
|
||
|
|
Use the bundled tokenizer and Qwen3 chat template. The output is a JSON array
|
||
|
|
with `entity`, `type`, `position`, `assertions`, and `candidates` fields.
|